Small Radiant Heat Floor Tankless Electric Hot Water - Help
Hello!
I installed a radiant heat floor in a 12'x12' solarium. All 4 walls are sliding glass doors. The floor is 3/4" wood. I ran 1/2" Oxygen Barrier PEX under the floor, stapled the PEX with Aluminum Heat Transfer, Radiant Barrier then Spray foamed insulation about 4". I used Hercules Cryo-Tek Corrosion Resistant Heating System Antifreeze in the system.
I have a Azel I-Link single zone controller with a Azel programmable Thermostat with the thermometer sunk into the wood floor about 1/2".
The water heater is a EcoSmart POU 6T 6.5kW 240v 30 Amp circuit. When I sized this out a 3.5kW would have been fine so I got the next step up as it was a nominal price increase. I know this heater is for point of use. I could not find one this small made for radiant heat. Open to others if they are out there.
Everything is working! Water is flowing, Thermostat is working. Water heater is set at 100 degrees.
I left the system running last night and the floor only gets to about 73-74 degrees. I am in the Boston Area.
Here is the issue. When I turn the water heater up past 100 degrees. It errors our ( E4) which means the incoming water is too hot.
When you look at the outgoing thermometer it's says it's about 110 degrees the incoming thermometer says its 109-110 degrees. When I take the temperature of the outside of the Grundfos motor it is about 115 degrees.
I AM SO CONFUSED. Why on earth is the water temp not dropping from the outgoing to the incoming? The heat should be dissipating and coming back in with at least a 10 degree drop. Why is the motor at 115 degrees? I feel like the system is generating heat itself. I cannot for the life of me figure out what is happening. Maybe this loop is too small to allow heat to escape?
Please do not roast me. This is a fun project. Any help is appreciated.
I want to get the floor to 80 degrees.
Comments
-
poor heat transfer to the floor, carpets? rugs?
is that a multi speed circ, set it to low(?),
known to beat dead horses0 -
Everything in the system has to balance. The heat released by the floor is determined by the difference between its surface temperature and the air in the room. That has to equal the heat delivered by the fluid, which is equal to the temperature delta in degrees F times the flow in GPM time 500.
If the temperature drop is not as big as you'd like, with the same heat output you need to reduce the flow. However, this will tend to reduce the temperature of the floor, which will reduce its heat output.
The question is going to be whether the water heater has a minimum output that is low enough to match what the floor actually puts out. If not, it may be easier to replace the tankless water heater with a small tanked water heater.
0 -
With the floor at 80F and the room at 72F the floor is going to produce 16 BTU/sf/hr. At 12x12 that's 144 square feet or 10,368 BTU/hr. That's just a hair over 3000 Watts. A point-of-use tank that takes 240V could easily produce that.
0 -
4 glass wall in a 12X12' room? Did you do a heatload calculation for the room?
Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream
2 -
@DCContrarian Are you suggesting use a small tank instead of the tankless? I do have a 30amp line running so that is easy todo.
0 -
Electro makes very small boilers. They can handle high incoming temps.
Which circulator is that? Looks like you have a tiny delta T between the two thermometers. Is that not what you expected? If the flow is quick enough, it’ll be small. Live and learn!0 -
3kW unit is about 10kBTU. to get a decent delta, you need at most about 1GPM through it.
That pump you have there with a simple 1/2" loop is good enough for probably 3-5GPM.Throttle the pump with a ball valve, you are way over pumping the setup.
1 -
Do you know what the heat loss of that room is? A quick back of the envelope calculation suggests to me that last night it may have been around 8,000 BTUh, unless those are very good windows. Hmmmm…
Now let's look at another number… you have about 5 gpm with that pump. You have a maximum heat loss through the floor of 10,000 BTUh. That should produce a delta T of around 3 degrees. You may not actually be far off that.
The pump is hotter, by the way, because it is water cooled — and the cooling water isn't cool…
So. Going back to the back of the envelope. You are probably putting out as much heat as that floor can put out, assuming that there is no carpet or furniture or what have you blocking part of it. That much heat isn't capable of getting the air temperature much higher, as the heat loss of the room is soaking it all up. The delta T of the water indicates that you are putting way too much water through the pipes…
Try reducing the flow through the floor, as has been suggested — no more than 1 gpm. That will give you a more manageable delta T. But… I suspect that you are going to struggle to get that space much more than 40 degrees or so warmer than it is outside, no matter what you do.
Br. Jamie, osb
Building superintendent/caretaker, 7200 sq. ft. historic house museum with dependencies in New England0 -
Another point: you may never see a 80F floor. You don’t get to pick that - it is dependent on your heat loss. If the thermostat shuts off at 76F, that’s what it is. If 80F is the temp for your heat loss, it won’t be today or tomorrow. It’ll be at 4am or so on the coldest day(s) of the year.
0 -
What is the heat load for the room?
Getting an 80° floor surface with wood, carpet and pad? That is going to be a tough go.
Are you wanting to maintain a room temperature or a floor surface temperature?
Slowing the flow the flow rate will not increase the heat output.
Here is a graphic of how that works. The tighter the delta, the higher the average temperature of the loop, the higher the output. Basic law of thermodynamics.
I suspect the carpet is the constipation. Pull apiece back, see what the wood surface temperature is.
Are you sure the loop(s) are purged and actually flowing?
Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream
2 -
@hot_rod I will pull back the carpet. The floor registers about 72 degrees when the water in the PEX is about 100ish.
Loops are purged. I ran them with a pump for 5-10 minutes into a 10 gallon bucket just letting it run. Closed one side of the loop got the pressure up to about 20 psi and shut off the other side. When it start to run I could hear the rest of the air vent from the vent. So I would assume they are all purged and running. Not sure how else to check. If you have any ideas happy todo so.
I would like to get the floor 70 - 80. I am not set on a temperature.0 -
This was a total finger blunder on the calculator by me. 16*144= 2304 BTU/hr. That's 675 Watts.
To get 3500 Watts — 12,000 BTU/hr — you would need 83 BTU per square foot. That would require a floor temperature about 42F above room temperature, or 114F. You'd probably have difficulty standing on that. You're going to need to add some sort of supplemental heat to the room, there's no way the floor is going to be able to deliver that. Most people find a floor above about 90F uncomfortable to stand on. At 90F the floor will deliver 36 BTU/sf or just over 5000 BTU/hr or 1500 Watts.
To get 675 Watts with a 10F temperature differential you'd need a flow of about 0.46 gallons per minute. With the observed 2F differential you'd need a flow of 2.3 GPM. That doesn't sound all that far off from what you're seeing.
1 -
Check for strainers on that heater, most have them to protect the valves and switches. It could be a restrictive heat exchanger. Getting 1.5 gpm through it with a circulator may not happen.
A small tank water heater may be better, a typical 4500W will give you about 15,000 btu/hr.
I think you may have a flow issue, and too much floor covering resistance.
Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream2 -
@Jamie Hall This makes a lot of sense. Thank you. This is a fun side project for an outdoor office I built years ago. Which I use year round. I have a mini-split HVAC that warms the air but the floor get's cold. I had extra pex from the house project I did years ago and figured it would be fun to see if I could get this to work just on a smaller scale. I am guessing the floor is not conducting enough heat.
You guys are all great. This gives me some ideas to try.0 -
@DCContrarian I have a Hyper Heat Mitsubishi Mini-Split that is the main heat. When I built this room I had extra pex from the house radiant heat floor project we did and installed it under this room. So my goal is to reasonably warm the floor, not as the main heat source, I'm not that crazy :)
Are you suggesting I slow the flow?
0 -
If you have the minisplit I think this simple setup will work for you:
Replace the tankless heater with a small (like 2.5 gallon ) point-of-use water heater. Set the thermostat on the water heater hot enough to get the floor to a comfortable temperature. Get a line-voltage thermostat like is used for electric baseboard heat and connect it to the circulator. So the circulator will only run when there is a call for heat, and the thermostat in the water heater will run the heating element as needed.
Most water heaters these days go up to 140F to fight legionnaires, I think that's going to be hot enough to get your floor to 80F.
Set the thermostat on the minisplit about 2F below the thermostat on the circulator. The minispit will only run when the floor heat is on and more heating is needed.
1 -
@DCContrarian Keep the pump at low with no other restrictions? Thank you for your time.
0 -
With the point-of-use tank the flow becomes much less important because it doesn't care what the return temperature is.
I guess on general principle you should have the flow at the lowest level that keeps the floor temperature where you want it. Higher flow uses more electricity* and will wear out your plumbing fittings marginally faster. But having the flow higher than optimal isn't going to cause any issues with the operation. Ultimately it depends on how easily the flow can be adjusted. I wouldn't buy a new pump just to fix the flow.
*(any electricity used by the pump is going to end up as heat in the room anyway so it's not really wasted).
0 -
one 1/2” pex loop? How long is it? That will be the determining factor on how much flow you can move, and how much heat you transfer. If you can eliminate the carpet you will feel the warm surface better. Good to know you have another source of heat
If that is a 15-58, on high speed it is around 78w. You may need speed 3 to over come the pressure drop in the loop and that heater. That wattage is the least of your cost if you have a 6 kw heater😉
With the loop or or loops length we can work backwards to an expected outcome.
Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream0 -
@hot_rod It's about 160' ish. 12'x12' = 9 joists; 8 spaces between each joist; about 20' of pex between each joist. The pump is a Grundfos 59896341 (UPS15-58FC). I had an extra lieing around so used it. The head from the pump to the floor is about 4'.
The heater I can send back still if needed. When I did the heat load it came to about 3Kw needed and I got the 6Kw as it was not much more. So if I need to switch that out for something else I can do that.0 -
While the heat load is 3.5kW, the floor itself is only going to be able to put out maybe 1500W max and more like 700W at your desired conditions. The heater you have is way oversized, and because of the tankless design it's not able to modulate that low. Definitely send it back. It doesn't do anything for you that a 1500W small tank heater wouldn't do.
0 -
If you get a 6 or 12 gallon electric tank, switch the element to a 5500W. That would give you 18,700 btu/hr.
However getting that through a single loop will be an issue. Here is what a 180' loop of 1/2' pex looks like at 1.5 gpm. Well beyond what that pump can move.
1 gpm at a 20∆ would get you 15,000btu/hr. So close to what a 4500W electri tank could provide.
That tankless heater depends on water pressure from a building to push water through 45- 60 psi or more. A small circ will not flow much through those small tankless.
I'd start with a tank heater, maybe a 30 or 40 gallon is cheaper, than the small 6 or 12, and it will be a 4500W element. Small tanks are generally 120V, 1600W elements.
Step 2 would be a hard surface floor covering. You can sneak up on a warmer surface, on a cold day with all that glass it may struggle.
Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream0 -
@ag27 wrote: "The water going into the floor is about 110 and coming out is like 108."
That really tells us everything we need to know about water flow. I have no idea how to predict the heat flow in this assembly, but I know that unless you modify the assembly the heat flow through it is going to be entirely determined by the difference between the room temperature and the average water temperature. Right now the average water temperature is (110+108)/2= 109F. With source water at 110F you're never going to get an average water temperature above 110F, which isn't going to give you meaningfully different heat flow from the current average of 109F. Increasing the flow might give you return temperatures of 109F, 109.5F, or even 109.9F, but none of them are going to meaningfully change the heat flow — and hence surface temperature — of the floor.
If I'm understanding @ag27 properly, the design goal isn't to provide a certain quantify of heat, it's to keep the floor at a certain temperature. At design temperature the floor is only going to provide about 15% of the heat for the room, the minisplit provides the rest. But heat flow and surface temperature are linked, surface temperature determines heat flow into the air. So if currently the surface temperature isn't what's wanted, the heat flow through the floor assembly needs to increase.
The heat flow through the floor assembly is entirely determined by three things: the average water temperature, the floor assembly construction, and the temperature of the room air. It's a little late now to change the floor assembly construction (although I consider rugs and furniture to be part of the assembly and they might be modified). And presumably you don't want to make the room colder just to make the floor warmer. That leaves water temperature. You need to increase the average temperature. That can be done by increasing the flow, or by increasing the source temperature. We've already looked at — and discarded — the flow. What needs to be done is increasing the source temperature.
One final note: the heat output of a floor like this is pretty sensitive to room temperature. In order to test the system you have to have room temperature at actual room temperature. In much of the east it was unseasonably warm for much of the past month, it was over 80F in DC last week. You can't test the floor under those conditions.
0 -
Electro has a 1.1kw boiler, maybe try that?
0 -
@DCContrarian Another option to change. I could modify the floor. I could put down and inch of concrete or other flooring on top of the 3/4" plywood subfloor to increase heat conductivity.
0 -
That would only increase the overall insulation level. More stuff=more insulation. The only way to get less insulation is to take stuff away.
0 -
@DCContrarian You are correct with the statement "the design goal isn't to provide a certain quantify of heat, it's to keep the floor at a certain temperature"
Would you recommend a small tank?0 -
Yes. A 2.5 gallon tank with a 1500 W element should be fine. Probably take up less space than what you've got there now.
0 -
The challenge with a system like this is temperature control
If the mini split maintains the room at say 70, and you want to feel a warm floor, say 80 f, then the floor adds heat at a rate of 2 btu/sq ft/ degree difference
So an 80 floor temperature in a 70 space will add 20 btus to the room and it will over-heat the space
I have tried this numerous ways for master bath floor warming, with dual stats and floor sensors, but never found a way to maintain a floor temperature and not over heat The floor needs to be at or above skin temperature to feel neutral or warm, around 80. Carpet tends to need to be even warmer to notice the warm toe feeling.
Ideally the floor would be the primary heating method, better yet the only heating method. But with all that glass, it may not be possible? The load calculation holds the key to this answer
A two stage thermostat with floor sensing limit might work out best. Let the floor do as much as possible, with an indoor/outdoor reset control, mini split as stage 2. So it does get complicated with two systems.
Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream0 -
I laid out my solution above:
Connect a line voltage thermostat to the circulator. Set it for 72F. Set the minisplit thermostat for 70F. Minisplit only kicks on when the floor isn't putting out enough heat to keep the room above 70F.
Note that from an energy usage perspective this is abysmal. The floor heat, COP 1.0, runs whenever any heating is needed. The minisplit, with a much higher COP, only runs when the floor isn't enough. In particular, in the mild weather where the minisplit is the most efficient it probably won't run at all. This is the opposite of what you'd do if you were optimizing for efficiency.
But the whole point of heating is comfort.
0 -
I may be thinking too far outside of the box here, but I like simple. 😁
Yours, Larry
3 -
Unless you are overwhelmed with info, here is an other option.
You mentioned glass sliders on all 4 walls? If so you will get some passive solar gain. Run the circ constantly. A Grundfos Alpha will run 20- 40W. So the solar energy that hits the floor will be circulated around the entire loop. I've done this on passive homes and on my shop currently.
Concrete or tile floors works best for this.
Then switch the heater on with a t-stat, via a RIB relay or fan center. @EdTheHeaterMan can draw that for you.
Do you have an interior wall to mount a t-stat? If so a single stats looks and works better. A two stage with floor sensor gives you more control options.
A pic of the room would help get an idea of the heat load and how much of the season that 1500W heater will cover the load without kicking on the mini split. You will enjoy the radiant much more.
Some temperature data for Boston. Assume the floor system could handle the heat load down to maybe 40- 45F. See where the mini would kick on
Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream1 -
A heated floor is going to put out 2BTU/hr for every degree F difference in temperature between the floor surface and the room temperature. With a surface temperature of 80F and a room temperature of 72F that's 16 BTU/hr per square foot; with 144 square feet that's 2304 BTU/hr.
@ag27 said his design load is 3500W and that he's in the Boston area. Design temperature for Boston is 13F (suburbs are probably a little lower). In BTU, 3500W is 12,000 BTU/hr. The design temperature is 59 degrees below room temperature. The floor will exactly balance when the indoor-outdoor temperature difference is 2304/12000= 0.192 of the design temperature difference, or 11F below room temperature, or 61F.
So when it's 61F or lower outside the floor can be kept at 80F, the heat pump runs to make up the difference, and everything is good. The question is, what happens when it's warmer than 61F? And the answer is another question, what do you want to happen? There's two alternatives: either you have the floor be a little bit cooler, or you let the indoor temperature creep up. For example, if you let the floor temperature drop to 78F, that's 6F above room temperature, 12 BTU/sf/hr, 1728 BTU/hr, 0.144 of design temp difference or 8F below room temp, which balances at 64F.
If you let the room temperature rise to 74F it's the same 6F difference and you end up with the same 8F below room temperature. But since the room is 74F now that balances at 66F.
If you do both, let the room rise to 74F and let the floor drop to 78F, that's 4F of difference, 8 BTU/hr/sf, 1152 BTU/hr, .096 of design temp difference or 6F below room temperature, balance at 68F.
My ideal would be to say at 75F outside the inside is at 75F, the floor is at 75F and the heat is off. From there as the temperature drops the indoor temperature drops, the floor temperature rises, until at 12F outdoor temperature the indoor temperature is at 72F and the floor temperature is at 80F. Below 61F outdoor the floor and indoor temperature stay constant.
Now, how practical is this? For a small system like this, not very. You need some way of sensing the outdoor temperature, and some way of modulating the water temperature and/or sensing the floor temperature. In high-end hydronics this kind of outdoor reset is very common but that's a whole other game. If you like to tinker, you could experiment with outdoor sensors and various ways of controlling the floor temperature.
If you don't want to tinker, the simplest control mechanism is modulating the floor temperature to keep the room temperature constant. This means that at 61F outdoor the floor will be at 80F and at 72F the floor will be at 72F, and at temperatures in between it will be a straight line. So at 67F outdoor the floor will be at 76F and the room will stay at 72F. In this scenario you don't measure or directly control the floor temperature, rather the thermostat turns the circulator pump on and off in response to the room temperature.
0 -
@DCContrarian @hot_rod First THANK YOU to both of you. I am a tinkerer, nerd and cybersecurity guy. So this is all great info. I have learned a ton. This is a fun project that I am learning with little consequence. FYI the wife wont kill me if I screw things up.
To answer @hot_rod I have a Azel Programmable Radiant Floor Thermostat mounted in side the room with a sensor in the plywood. Between each of the sliding glass doors there is about a 6" flat wood spacer.
Also @hot_rod and @DCContrarian The floor would not be on unless it drops below 45-50. the issue I have is in the winter and it's 10-20 degrees out, the floor is freezing. The mini-split is fine for the air but the floor is freezing. We did insulate it very well, but there is no heat down there and it's raised.
I love the idea of sensing outdoor temps to adjust the floor as needed. If it is not overly complicated.
Here is a picture:
0 -
A great looking space!
I'd figure close to 400 sq feet of wall at maybe R-5 with the glass an lumber.Doesn't look like much wall framing to insulate. What's the r value in the ceiling and floor? It just seems the loads would be more than 27 btu/ft in a space like that?
It's going to come down to how much $$ and technology you want to throw at it. The small boilers like the Electro MN and Thermo 2000 have outdoor reset built into some of their models, and they modulate.
I used one of these Thermo 2000 mini's in a 160 sq ft tiny home I build. I put a lower wattage element in it to avoid short cycling. I think the smallest they offer is 3 kw.
This gives you a real, low pressure drop electric boiler with the outdoor reset onboard. Plus more KW/btu if you can drive the floor higher. If the carpet stays you might run 85F?
If you can get in the crawl space this JAGA "canal" heaters could run off the boiler and give you the boost you need without the split running. They have some cool corner radiators also.
Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream0 -
It sounds like more of a cold floor issue than need for warm toes feeling. There is a big difference between the two and cold floor is much easier to deal with.
Here is what I've done that works. Get a floor heat thermostat that can do air and floor temp control (I've used Honeywell TH115-AF before for this). Configure the minimum floor temp to something like 75F and the air temp target a bit bellow your heat pump setpoint.
This will fix the cold floor issue while the bulk of your space heat would still come from the heat pump. The floor heat would only run enough to keep the slab temp up but not enough to overheat the place.
0 -
I'd like to see 100% of the heating be hydronic.
When we have rooms that don't have enough floor square footage to cover the heat load, we add supplemental heat. This is common in kitchens where the cabinets eat up so much of the floor space. The floor is a radiant heat emitter. Being low temperature you need a lot of surface area.
Kick space heaters or some sort of panel rads, are often used as supplemental heat. . Jaga makes heat emitters that run at radiant temperatures for that extra output. That simplifies piping as you need just one temperature.
Anything you can do to keep that heat load below the floor lower would help. Is there perimeter insulation around the foundation? That is a big cold spot. The lower the heat load the easier and less $$ to heat or warm the floors.
Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream0
Categories
- All Categories
- 86.2K THE MAIN WALL
- 3.1K A-C, Heat Pumps & Refrigeration
- 52 Biomass
- 422 Carbon Monoxide Awareness
- 90 Chimneys & Flues
- 2K Domestic Hot Water
- 5.4K Gas Heating
- 99 Geothermal
- 156 Indoor-Air Quality
- 3.4K Oil Heating
- 63 Pipe Deterioration
- 912 Plumbing
- 6K Radiant Heating
- 380 Solar
- 14.8K Strictly Steam
- 3.3K Thermostats and Controls
- 53 Water Quality
- 41 Industry Classes
- 47 Job Opportunities
- 17 Recall Announcements